The overall goal of this project is to test the hypothesis that immunization against antigens expressed by tumor cells can result in meaningful anti-tumor effects in patients. Work is focussed on two types of antigens. The first type of antigen, represented by GD3 ganglioside, is a differentiation antigen - a normal molecule expressed by cells as part of the differentiation program. We have developed 2 vaccine constructs that can induce anti-GD3 antibodies in patients: BEC2 anti-idiotypic monoclonal antibody combined with BCG adjuvant, and GD3-lactone-KLH combined with QS21 adjuvant. The specific aims for the GD3 system are: a) To determine the relative immunogenicity of BEC2/BCG vaccine at a range of BEC2 doses; b) To determine whether prolonged immunization with BEC2 boosts the anti-GD3 antibody response; c) To assess whether priming with BEC2/BCG will enhance the anti-GD3 antibody response induced by immunization with GD3-lactone- KLH/QS21; d) To assess whether priming with GD3-lactone-KLH/QS21 will enhance the anti-GD3 antibody response induced by immunization with BEC2/BCG; e) In limited disease small-cell lung cancer, determine whether immunization with BEC2/BCG following initial chemoradiotherapy will enhance overall survival. The second type of antigen -- mutated oncogenes -- represent absolutely tumor-specific targets. As a model for this type of antigenic target, we are carrying out vaccine studies to immunize against mutated K-ras protein in pancreatic carcinoma. Specific aims for the K-ras system: a) To test the hypothesis that immunization against the K-ras mutation present in a patient's tumor can induce a specific T cell response against the K-ras mutation; b) To observe whether skin test reactivity or HLA type correlates with the induction of anti-K-ras T cell response; c) To observe any anti-tumor effects. To accomplish these specific aims, we propose a series of 4 clinical trials. The other objective of this grant proposal is to provide an opportunity to train beginning clinical investigators. This will be accomplished by having Medical Oncology fellows participate in every aspect of performing the trials, analyzing the data, and presenting the results. This close mentoring relationship reflects our current teaching paradigm. The grant award will provide the PI with more time to conduct the clinical trials and to mentor additional beginning clinical investigators.